Data analysis program

Programs like Microsoft Excel can be used for graphing data, but graphing with LabVIEW is often faster, especially when large amounts of data are dealt with.

Microsoft Excel 2007 is limited to 32000 data points per series in a chart (Microsoft Corporation, 2011), and the processing time can be extensive. Although LabVIEW can provide faster graphing, some programming is required. However, this can be advantageous since the program can be adapted to user requirements.

A data analysis program was written in LabVIEW for rapid data viewing and analysis. There are four sections to the data analysis program: Open File, Signal Analysis, Save Filtered Profile, and File Data tabs. The following sections describe the operations and features of each in detail.

2.1.2.2.1 Open file tab

Before running the data analysis program a file path must be selected. Since most shots are saved by shot number in a common directory, a file path can be selected based entirely on the shot number. Selecting a particular shot is done by using the "Shot Number" selection, and running the program. The sampling frequency, number of channels, and the number of samples per channel are automatically determined from the file and displayed. The program can be stopped at any time using the "STOP" button located on the left side of the user interface.

Figure 2-7: Data analysis program: "Open File" tab

If the desired file is not named after a particular shot, the "Manual Select File" button must be turned off. The URL entered in the file path will be opened. The sampling frequency may not be detected when manually selecting a file, so the "Manual Select Sample Rate" button can be used to enter the appropriate sampling rate.

2.1.2.2.2 Signal analysis tab

The signal analysis tab has a variety of options which were developed specifically for analyzing shock wave profiles, but can be implemented for other purposes as well.

There are two graphs showing the data in a time domain, and a third which shows the Fourier transform (FFT) of a given profile in the frequency domain.

The first graph shown in Figure 2-8 shows a single profile which is selected using the "Select Profile" indexer. The start time and the duration are adjusted using the "Start Time" and "Graph Duration" indexers. The adjustment increments can be set using the

"Start Increment" and "Duration Increment" indexers. For example, an incremental rate set at -3 adjusts the time interval by 0.001 (10^-3) seconds. When the data were collected, an approximate arrival time was determined at 25% of the peak signal value. Pressing the "Go to Arrival Time" button will shift the start of the graph to this time allowing the user to adjust a graph to the shock front quickly. This feature works consistently for rapid rising profiles, but is not as effective for slow rising signals.

The impulse under a curve is automatically calculated and displayed in the "Impulse"

output box. The calculated impulse is graphically represented by the area under the first graph in Figure 2-8 with the negative regions subtracted. When calculating the peak overpressure of a shock wave the average value of several points is used, especially for sensor calibration using flat-topped shock waves. The "# Points Averaged" selection adjusts the number of points to average after the start time of the first graph. The average is shown in the "Average Value" display box, and a cursor graphically indicates the average value and the duration of the averaging.

Figure 2-8: Data analysis program: Signal Analysis tab

The second graph in Figure 2-8 shows multiple waveforms simultaneously.

Displayed waveforms can be selected using the "Visible Waveforms" array of switches, which simultaneously adjusts the legend on the right of the graph. The scaling of the horizontal axis of this graph is scaled identically to graph 1, but the vertical axis is scaled automatically for all signals. When comparing waveforms on this graph, it is sometimes useful to compare signals to a given value. For this reason the cursor on graph 2 can be adjusted with the "Vertical Cursor Position" control, which can be automatically set to the average value computed for graph 1.

Graph 3 shows the Fourier transform of the selected profile in graph 1 and is normalized to have a peak value at unity. The frequency range can be adjusted from zero to half of the sampling frequency using the "Frequency Maximum" and "Frequency Minimum" controls. The frequency increment between points is equal to the sampling frequency divided by the number of samples taken. This analysis tool is useful for observing the main frequencies present in a signal. The largest frequency peak of a shock wave is at a low frequency, but minor spikes can also show high frequency components, noise, and resonant frequencies of sensors.

Undesired high frequency noise based on sensor dynamics does not necessarily represent part of an actual shock wave signal, and can be reduced using filtering techniques. A filtering algorithm was implemented to remove unrepresentative frequencies (noise). Several filter types are available including the low-pass, high-pass, and band-pass, but the low-pass filter is used almost exclusively. Filtering can be turned on or off using the "Filter Signal" button. The filter characteristics (type, order, and cutoff frequencies) can be selected using the controls below the “Filter Signal” button.

Using a low-pass filter with a cutoff of 25 kHz cleans up a shock wave profile significantly, with less than 0.1% decrease in calculated impulse.

2.1.2.2.3 Save filtered profile tab

Filtered shock wave profiles can be successfully used as inputs to finite element simulations such as ABAQUS, without introducing unrealistic noise created by the imperfect pressure sensor dynamics. Filtering of a signal does not change the data in the original file, so the capability to save filtered data was created. The "Select Profile"

control in the "Signal Analysis" tab determines which filtered profile will be saved. The

"Write Filtered Signal File" should be pressed and held until a prompt opens to enter a file path and name. Clicking "OK" saves the filtered profile in a separate text file which can be imported into other programs.

Figure 2-9: Data analysis program: Save Filtered Profile tab

2.1.2.2.4 File data tab

The “File Data” tab contains an array with data from the opened text file in tab delimited format. The starting points of the array can be adjusted using the arrows on the upper right of the array. All of the user inputs are contained in this array, as well as the approximate arrival time, DC offset before shifting, and the shot data (pressure, strain, etc.). An example showing the information in this tab is shown in Figure 2-10.

Figure 2-10: Data analysis program: File Data tab